John Griffiths
Sociology of Law, Faculty of Law, University of Groningen, Groningen

2. An empirical concept of law

One cannot look for the poles of the Earth until one knows what a pole is – that is, what the expression ‘pole’ means, and also what counts as finding a pole of the Earth. Otherwise, like Winnie-the-Pooh, one might embark on an expedition to the East Pole.2828. M. Bennett and P. Hacker, Philosophical Foundations of Neuroscience, Malden: Blackwell, 2003, 71 (with thanks to Henri Wijsbek for calling this to my attention).View all notes

The first problem for the sociology of law … is to identify the sort of social fact it takes as its subject matter.2929. Griffiths (1984aGriffiths, J. 1984a. “The Division of Labor in Social Control.” In Toward a General Theory of Social Control, edited by D. Black3770Orlando, FL: Academic Press.[Crossref], [Google Scholar], 39).View all notes

The concept “law” denotes – on first impression – the object of the sociology of law. Nevertheless, after well over a century of attention to the matter, the definition of “law” as an object of empirical research and theory – that is, an empirical concept of law – is still fraught with difficulty and confusion.

2.1. The various manifestations of “law”

A quick look at what Galanter has called the “manifestations of law” – that is, things commonly referred to as “law” or “legal” – shows how various the phenomena are that fall within everyday usage of the word “law” (see Box 1). “Law” (droit, Recht, derecho, ius) is used to refer, more or less indiscriminately, to a whole variety of phenomena that seem related in some way to one another, but which certainly do not amount to the same thing. There are probably not many statements that are true for all these various “manifestations of law”. So if someone asserts something about “law” (such as its fairness or its effectiveness), then the first question one needs to ask is: which manifestation is he or she talking about? The effectiveness of legal education in the formation of lawyers is surely quite a different matter from the effectiveness of the legal institution of property in bringing about the efficient use of resources. And there is no point to criticism of Black’s “theory of law” – in which “law” is conceived of as the social control behavior of government officials – on the ground that it does not permit us to explain the contents or effects of legal rules.

Galanter’s overview makes one view with suspicion any claim that one or another of law’s manifestations has a privileged position with respect to the concept of law. For a long time, the idea that law is essentially a collection of rules (“commands of the sovereign”, for example3030. This is J. Austin’s ((1832Austin, J. (1832) 1965. The Province of Jurisprudence Determined. Reprint, LondonWeidenfeld and Nicolson.[Crossref], [Google Scholar]) 1965) classic definition of law, on which many others – including those of Malinowski and Pospisil (see Box 2) – are variations. Cf. Hart (1961Hart, H. 1961The Concept of Law. (2nd ed. 1994). OxfordOxford University Press. Page references in this essay are to the 1961 edition. [Google Scholar], 25).View all notes) was widely accepted.3131. See Ehrlich, Malinowski, Hoebel, Bohannan, Pospisil (Box 2).View all notes Alternative conceptions see law as social order, maintained in a particular way, or a particular form of social control.3232. See Weber, Radcliffe-Brown, Pound, Schwartz, Selznick, Black (Box 2).View all notes A few decades ago, the “dispute” paradigm of law was prevalent,3333. See Pospisil (Box 2). Llewellyn and Hoebel (1941Llewellyn, K., and E. Hoebel1941The Cheyenne Way: Conflict and Case Law in Primitive JurisprudenceNormanOklahoma University Press. [Google Scholar]) were very influential in promoting this conception of law, especially among anthropologists (cf. also Abel 1973b).View all notes until Holleman put an end to the idea that law manifests itself exclusively or even primarily in “trouble cases” by calling attention to the importance of “trouble-less cases”: the use of law as a “language of interaction” in everyday transactions.3434. Holleman (1973Holleman, J. 1973. “Trouble-Cases and Trouble-Less Cases in the Study of Customary Law and Legal Reform.” Law and Society Review 7: 585609.[Crossref], [Google Scholar]). The expression “language of interaction” is taken from Fuller (1969Fuller, L. 1969. “Human Interaction and the Law.” The American Journal of Jurisprudence 14: 136.[Crossref], [Google Scholar]). Ehrlich’s (1936Ehrlich, E. 1936Fundamental Principles of the Sociology of Law. Translated by W. L. MollCambridgeHarvard University Press. [Google Scholar]) sociology of law shares this insight.View all notes

In short: without some clarification of the concept “law”, it is impossible to say anything sensible at all about what the sociology of law is.

Box 1: Galanter’s “manifestations of law”

• legal norms, which include:

legal rules, among which one can distinguish:

legislation: rules adopted through a formal procedure and having an authoritative formulation
common law: rules formulated by judges in the course of deciding cases
customary law: rules that emerge in social practice and are commonly regarded as binding
legal principles and values, such as the principle that no one can be a judge in his own case, and the value of legal certainty
legal policy, such as the importance of efficiency in the administration of justice
  • legal institutions such as property, family, contract

  • legal organization: the legal profession(s), adjudicatory, legislative and administrative bodies, the police and so forth

  • legal processes: rule-making, judging, enforcement, administration, execution

  • legal reasoning: arguments generally accepted among lawyers as good reasons for accepting R as a legal rule or for interpreting R in a particular way

  • legal behavior/interaction: behavior/interaction ordered by the use of legal rules

  • legal scholarship and legal education

  • the legal system: all of the above taken as a whole (whereby it must be noted that the degree of coherence and autonomy of such a system can vary greatly)

  • Source: derived with a few minor changes from Galanter (1977Galanter, M. 1977. “Notes on the Future of Social Research in Law.”. In Law and the Behavioral Sciences, edited by L. Friedman and S.Macaulay1820New YorkBobbs Merrill. (1974, unpublished paper). [Google Scholar]).

2.2. What’s in a name?

What’s in a name? that which we call a rose, by any other name would smell as sweet.3535. W. Shakespeare, Romeo and Juliet, Act II, scene 2. This is the functional approach to the concept of law: the term “law” refers to whatever it is that is involved in one of what Llewellyn called the “law jobs” (see Twining 1985Twining, W. (1973) 1985Karl Llewellyn and the Realist MovementLondonWeidenfeld & Nicolson. [Google Scholar], 175ff).View all notes

Law is only law if it is labeled ‘Law’.3636. Pieterman (1993b, 132) (“recht is alleen recht als er recht op staat”), paraphrasing an old Dutch candy advertisement that amounts to about the same as “Coke is only Coke if it says ‘Coke’ on the label”. This is the Baron von Munchhausen [legal positivist] approach to the concept of law, named after the mythological character who sought to save himself from sinking into a bog by pulling himself up by his hair. It is the normal approach of legal scholars and uncritical sociologists of law.View all notes

When I use a word, it means just what I choose it to mean – neither more nor less.3737. L. Carroll, Through the Looking Glass, and What Alice Found There.London: Macmillan 1871, Chapter 6. This saying of Humpty Dumpty is the Frank Sinatra approach to defining the concept of law: everyone does it his way.View all notes

Rose is a rose is a rose is a rose.3838. G. Stein, in Bartlett’s Familiar Quotations, 14e ed., 933a. This is the essentialist approach to the concept of law: it is what it is (and you know it when you see it).View all notes

Ceci, ce n’est pas une pipe.3939. This is the title that is part of Magritte’s painting of a pipe. It might be called the natural law approach to the concept of law (“Nazi law is not law”).View all notes

2.2.1. Unsatisfactory approaches to an empirical concept of law

The common element in Galanter’s “manifestations of law” is the adjective “legal”. Whether their focus is on individual rules or, at the other end of the spectrum, on whole legal systems, what various authors seek to capture in the concept “law” is that property that distinguishes one or another of the manifestations of law from other similar – but not “legal” – phenomena. Those whose perspective is internal find this difficult, but not impossible. Lawyers construct the concept “law” in the course of their professional behavior. Hart (1961Hart, H. 1961The Concept of Law. (2nd ed. 1994). OxfordOxford University Press. Page references in this essay are to the 1961 edition. [Google Scholar]) describes what they do, when they identify some phenomena as “legal” and others as “not legal”, as the application of a “rule of recognition”; in any given society the validity of that rule, according to Hart, is a matter of fact.4040. Hart (1961Hart, H. 1961The Concept of Law. (2nd ed. 1994). OxfordOxford University Press. Page references in this essay are to the 1961 edition. [Google Scholar], 107).View all notes So long as most lawyers, most of the time, identify the “legal” in more or less the same way – that is, so long as a society’s legal order is not fundamentally challenged – there is no problem. But for anyone in search of a secure external (empirical) basis for identifying the distinctively “legal”, Hart’s analysis has a distressingly Baron-von-Münchausen quality.4141. See note 36.View all notes Hart himself observed that his philosophical analysis of the concept of law is ultimately “an essay in descriptive sociology”: law is what the members of a given social group call and accept as law.4242. Hart (1961Hart, H. 1961The Concept of Law. (2nd ed. 1994). OxfordOxford University Press. Page references in this essay are to the 1961 edition. [Google Scholar], vii).View all notes Ironically, this is precisely what makes his concept of law unsuitable for the sociology of law. Consensus within a group about what its “law” is, identifies for the external observer something that is apparently an important normative category in that group, but not a category that is usable for external, empirical and hence (necessarily) comparative purposes.4343. For this reason, among others (such as the latent terminological circularity it shares with Hart’s definition and its inapplicability in many cases of what is generally regarded as legal pluralism, cf. n 52), Tamanaha’s suggestion (2001Tamanaha, B. 2001A General Jurisprudence of Law and SocietyNew YorkOxford University Press.[Crossref], [Google Scholar], 167) that a social scientist interested in legal pluralism should regard as “law” in any social group that which “sufficient people with sufficient conviction consider … to be ‘law’…” is unusable for purposes of sociology of law.View all notes

One can, of course, define a word any way one chooses. When Black chooses to define “law” as “governmental social control”,4444. Black (1976Black, D. 1976. The Behavior of Law. New YorkAcademic Press. [Google Scholar]).View all notes it would be pedantic to object that the Oxford English Dictionary, lawyers, Joe the Plumber, Herbert Hart, or anyone or everyone else uses the word in some different way. Nevertheless, some definitions will be more useful – for some purposes – than others. I have come to share the view that because of the associations that the word “law” (and its derivatives, such as “legal”) share with particular ideas about how social order ought to be maintained, it is in practice impossible to purge the word “law” by means of a mere definition of its normative connotations.4545. For an example of such an approach, see Griffiths (1984aGriffiths, J. 1984a. “The Division of Labor in Social Control.” In Toward a General Theory of Social Control, edited by D. Black3770Orlando, FL: Academic Press.[Crossref], [Google Scholar]), justly criticized for this by van den Bergh (1986van den Bergh, G. 1986. “Over een theorie van sociale controle en het meten van recht [Concerning a Theory of Social Control and the Measurement of Law].” Recht en Kritiek 12: 374386. [Google Scholar], 380–382).View all notes To borrow an expression from the philosopher J. L. Austin, “law” (and its relatives in other languages – droitRecht and so forth) is a word that comes “trailing clouds of etymology”4646. Austin (1961Austin, J. L. 1961. “A Plea for Excuses.” In Philosophical Papers, edited by J. O.Urmson and G. J. Warnock123152OxfordClarendon Press. Reprinted from Proceedings of the Aristotelian Society 19561957. [Google Scholar], 149).View all notes and whatever one does by way of definition, these clouds will forever get in the way of clear thought. I conclude that the term “law” should not be used to denote the key variable in the sociology of law.4747. Abel (1973bAbel, R. 1973b. “A Comparative Theory of Dispute Institutions in Society.” Law and Society Review 8 (2): 217347.[Crossref], [Google Scholar]) is one of the few who have seriously confronted the question whether “law” is a suitable theoretical concept in the sociology of law. His pessimistic conclusion anticipates the conclusion reached here (224): “[f]or the time being, at least, it seems clear that we must displace law from the center of our conceptual focus as we attempt to build social theory”.View all notes If we want to develop social scientific theory about “law” we must look for a more suitable theoretical concept.

2.2.2. The search for an empirical concept of “law”

A quick look at the history of efforts to define “law”’ for purposes of empirical study is chastening. The search for an empirical concept of law – one not dependent on the normative preoccupations of lawyers and other participants but rather on the demands of sociological theory – has so far proven unproductive.4848. Compare Moore (1969bMoore S. 1969b. “Law and Anthropology.” In Biennial Review of Anthropology, edited by B. J. SiegelStanfordStanford University Press. Reprinted in Moore 1978. London: Routledge & Kegan Paul. [Google Scholar]) for an extensive overview of the anthropological literature on the concept of law; see also F. and K. von Benda-Beckmann (2006von Benda-Beckmann, F. and K. von Benda-Beckmann2006. “The Dynamics of Change and Continuity in Plural Legal Orders.” The Journal of Legal Pluralism and Unofficial Law 38: 144.[Taylor & Francis Online], [Google Scholar]).View all notes Box 2 collects a small chronological sample of some of the most influential proposals (the preponderance of anthropologists of law in the sample reflects the fact that they have always been more aware than sociologists of law of the necessity of a general definition – that is, one suitable for intercultural and historical comparison).

Box 2: Proposed empirical concepts of law

Ehrlich (1936Ehrlich, E. 1936Fundamental Principles of the Sociology of Law. Translated by W. L. MollCambridgeHarvard University Press. [Google Scholar], 24)

A rule of law is an “ordering” of the social relationships in a group (“association”): “a rule which assigns to each and every member of the association his position in the community…and his duties.”

Weber (1954Weber, M. 1954On Law in Economy and Society. Translated by Rheinstein and Shils. New YorkSimon and Schuster. (1st German edition. 1922). [Google Scholar], 5)

Law is order that is “externally guaranteed by the probability that coercion (physical or psychological), to bring about conformity or avenge violation, will be applied by a staff of people holding themselves specially ready for that purpose.”

Malinowski (1926Malinowski, B. 1926Crime and Custom in Savage SocietyLondonRoutledge. [Google Scholar], 55)

“The rules of law stand out from the rest in that they are felt and regarded as the obligations of one person and the rightful claims of another. They are sanctioned not by a mere psychological motive, but by a definite social machinery of binding force….”

Radcliffe-Brown (1965Radcliffe-Brown, A. 1965. “Primitive Law.” In Structure and Function in Primitive Society, edited by A .R.Radcliffe-Brown212219New YorkFree Press. [Google Scholar], 212, quoting Pound)

Law is “social control through the systematic application of the force of politically organized society”.

Llewellyn and Hoebel (1941Llewellyn, K., and E. Hoebel1941The Cheyenne Way: Conflict and Case Law in Primitive JurisprudenceNormanOklahoma University Press. [Google Scholar], 23)

“[T]he legal is best seen as that which is marked by authority – which is recognized as imperative…”.

Pound (1942Pound, R.1942. Social Control through Law. New Haven: Yale University Press. [Google Scholar], 25)

Law is social control by the state – “the systematic and orderly application of force by the appointed agents”.

Hoebel (1954Hoebel, E. A. 1954The Law of Primitive ManCambridge, MAHarvard University Press. [Google Scholar], 28)

“A social norm is legal if its neglect or infraction is regularly met, in threat or in fact, by the application of physical force by an individual or group possessing the socially recognized privilege of so acting.”

Schwartz (1954Schwartz, R. 1954. “Social Factors in the Development of Legal Control: A Study Case Study of Two Israeli Settlements.” The Yale Law Journal 63: 471491.[Crossref], [Web of Science ®], [Google Scholar], 473)

Legal control is “social control … which is carried out by specialized functionaries who are socially delegated the task of intra-group control…”.

Bohannan (1965aBohannan, P. 1965a. “The Differing Realms of the Law.” In The Ethnography of Law, edited by L.Nader3342. Special Publication. American Anthropologist 67 (7), part 2.[Crossref], [Google Scholar], 36)

“Law is … ‘a body of binding obligations regarded as right by one party and acknowledged as the duty by the other’ (quoting, with several minor errors, Malinowski [1926Malinowski, B. 1926Crime and Custom in Savage SocietyLondonRoutledge. [Google Scholar], 58]) which has been reinstitutionalized within the legal institution so that society can continue to function in an orderly manner on the basis of rules so maintained”. (Italics in original)

Selznick (1969Selznick, P.P. Nonet, and Howard M.Vollmer1969. Law, Society and Industrial Justice. New York, NYRussell Sage Foundation. [Google Scholar], 7)

Law is “endemic in all institutions that rely for social control on formal authority and rule-making”. (Italics in original)

Pospisil (1971Pospisil, L. 1971Anthropology of Law. A Comparative PerspectiveNew YorkHarper and Row. [Google Scholar], 95, 56)

Law consists of “principles of institutionalized social control, abstracted from decisions passed by a legal authority (i.e. a ‘person whose decisions are followed’)”.

Black (1976Black, D. 1976. The Behavior of Law. New YorkAcademic Press. [Google Scholar], 2)

“Law is governmental social control.”

Despite their superficial differences, the definitions quoted share a common structure. Social scientists commonly consider law to be a special case of a more fundamental and general social phenomenon (morality, social control, social order, politics, rules). The more specific category “law” is defined by identifying it as a subcategory of the more general category. I refer to all such candidates for an empirical concept of law as “taxonomic”, by way of analogy to the systematic way different species are identified in biology. The structure of such definitions of law is shown in Figure 1.

Figure 1. The structure of a taxonomic concept of law.

Among social scientists, the most common taxonomic concept of law considers it to be a special case of “social control”: “specialized” social control (Schwartz), “governmental” social control (Black), “authoritative” social control (Selznick), “reinstitutionalized” social control (Bohannan) and so forth. In what follows, I limit myself to taxonomic definitions of “law” as a special case of “social control”.4949. Other taxonomic conceptions of “law” see it as a special case of commands (Austin), rules (Hart, Malinowski), politics (cf. Gulliver 1963Gulliver, P. 1963Social Control in an African SocietyLondonRoutledge & Kegan Paul. [Google Scholar], 299) or positive morality (“social norms”: Hoebel). For F. and K. von Benda-Beckmann (2006von Benda-Beckmann, F. and K. von Benda-Beckmann2006. “The Dynamics of Change and Continuity in Plural Legal Orders.” The Journal of Legal Pluralism and Unofficial Law 38: 144.[Taylor & Francis Online], [Google Scholar], 12), “law” is a special case of “objectified cognitive and normative conceptions for which validity…is authoritatively asserted”.View all notes

Schwartz’ definition of “legal control” is one of the most thoughtful in the whole taxonomic tradition and its defects are characteristic of such an approach. It, therefore, deserves special attention. His study concerned two agricultural settlements in Israel whose social control practices were strikingly different. He describes the differences he observed in terms of the dichotomy “legal” and “informal”. “Legal” control he defines as social control “delegated to specialized functionaries”. One of the two settlements (the moshav) had “legal control” in this sense: a “Judicial Committee” charged with hearing and deciding conflicts between members. The other settlement (the kvutza) had no such institution but relied entirely on “informal” control. Of course, from the point of view of the Israeli state, neither practice was “legal”.5050. There is also no reason to suppose that either group of settlers would have regarded the social control in their settlement as “law” – they probably reserved that term for the “law” of the Israeli state. But to be clear about the matter: paceTamanaha (see n. 43), their “internal” point of view on the matter is not relevant as far as an empirical concept of law is concerned.View all notesBut Schwartz obviously felt the need to define “legal” in a different way for social scientific purposes.

Schwartz uses the term “specialization” to describe the Judicial Committee, but as far as one can tell from his description, the members of the Judicial Committee in the moshav were not “specialists” in social control: their special training, skills and experience had to do with farming, not with adjudication.5151. The formulation of the ‘relational distance proposition’ in Griffiths (1984aGriffiths, J. 1984a. “The Division of Labor in Social Control.” In Toward a General Theory of Social Control, edited by D. Black3770Orlando, FL: Academic Press.[Crossref], [Google Scholar]) (discussing Schwartz 1954Schwartz, R. 1954. “Social Factors in the Development of Legal Control: A Study Case Study of Two Israeli Settlements.” The Yale Law Journal 63: 471491.[Crossref], [Web of Science ®], [Google Scholar]) was unfortunate in using the term “specialization”.View all notes Nor did they perform some special part of a whole task (such as the making of pins, in Adam Smith’s famous example of the increased productivity attendant upon the “division of labor”5252. Smith (1776Smith, A. 1976An Inquiry into the Nature and Causes of the Wealth of NationsNew YorkRandom House. [Google Scholar], Book 1, Ch. 1).View all notes). In short, they were not specialized but differentiated functionaries: set apart – when engaged in adjudication rather – from the rest of the members of the settlement. The adjudicatory function was exclusively theirs, but it was one even they could not perform in the course of everyday life but only at special times and places and according to specific procedural rules. It was thus a function differentiated from the rest of social life in the settlement.5353. It is also not the case that specialization implies differentiation. One can be “specialized” in something without necessarily occupying a distinct social role to which performance of that thing is allocated: a particular sort of medical “specialist” such as a surgeon may or may not be entitled to practice (a particular variety of) surgery to the exclusion of all other doctors.View all notes I will henceforth use the term “differentiated” social control in reconstructing Schwartz’ argument and setting out my own conception of “law” for purposes of the sociology of law.

Differentiation: Although the idea of differentiation in social control is frequently invoked in the sociology of law, I know only one serious and explicit treatment of the concept: Abel’s (1973bAbel, R. 1973b. “A Comparative Theory of Dispute Institutions in Society.” Law and Society Review 8 (2): 217347.[Crossref], [Google Scholar]) article on dispute institutions in society. Abel regards “structural differentiation” as a key explanatory variable for variation in dispute processes (see, in particular, 253–262). In particular, his variable “role differentiation” – “defined by explicit prescriptions” that “demarcate private life from official business”, “circumscribe the powers of the intervener” and “regulate conduct within the dispute institution” (262) – comes close to the definition I propose here.

I define differentiation as follows: the existence in a social group of secondary rules creating distinct social roles for the performance of particular social tasks (cf n 62 and §3.2.2). For example, social control in a school in which the maintenance of order in the hallways during breaks is not carried out by the teachers in the normal course of their daily activities, but is recognized as a distinct role and allocated to security guards, is more differentiated than is social control in a school where this is not the case. (I owe this example to a newspaper article on the budgetary difficulties of New York City public schools resulting from increased school size and the attendant reliance on hired guards to maintain order.)54

For purposes of describing social control in the two settlements and of the explanatory theory Schwarz brings to bear,5555. Schwartz’ formulation of the theory is in embarrassingly “functional” terms – societies get the sort of control they “need” – but nevertheless, after more than half a century, his analysis is still one of the most sophisticated to be found in the literature.View all notes a rough-and-ready dichotomy between “differentiated” and “undifferentiated” social control was perhaps sufficient. Nevertheless, the closer one thinks about the matter, the clearer it becomes that it is simply not true that differentiation in social control was only present in one of the two settlements. While perhaps less clearly visible than in the case of the moshav’s Judicial Committee, differentiated social control was also present in the kvutza. For example, a committee whose principal task was organizational engaged, as Schwartz shows, in social control as well. But most importantly, the kvutza had a “General Assembly” which occasionally imposed the most drastic sanction known to the community: expulsion. Although almost all adult inhabitants were members, social control by the General Assembly was differentiated: like the farmers who constituted the Judicial Committee of the moshav, the members could only engage in such social control while in a socially regulated “official” capacity as members of the General Assembly, they could not do so on their own in the course of everyday interaction. If we take Schwartz’ definition of “legal control” literally, there was thus “legal control” in both settlements. The difference between them that he sought to explain did not, by definition, exist.

Schwartz was trapped, in effect, in the fundamental point of departure of the whole taxonomic approach, namely that the criterion that serves to define “law” as a subset of social control knows only two values. But differentiation is a continuous variable. If one focuses on that rather obvious point, the whole taxonomic tradition in sociology of law begins to seem a little perverse. The other taxonomic criteria that have been proposed (“authority”, “institutions”, even the “state”) all share the same problem: one cannot use them to divide social control into two sorts (legal and non-legal) because they are not dichotomous but continuous variables.

Further objections to taxonomic concepts of “law” include the following:

  • The continuum of increasing differentiation knows no inherent discontinuities (such as the freezing point of water) of which one can say: past this point, all social control is “legal”. The choice of a given point is empirically arbitrary. What the choice generally seems to reflect is an ethnocentric preference for what is called “law” in one’s own culture.5656. Cf. Abel (1973bAbel, R. 1973b. “A Comparative Theory of Dispute Institutions in Society.” Law and Society Review 8 (2): 217347.[Crossref], [Google Scholar], 222–224).View all notes

  • All taxonomic concepts of law are both under- and over-inclusive. Observed social control phenomena on opposite sides of the taxonomic criterion (see Figure 1) but otherwise close to each other, will often have far more in common than either has with phenomena on its own side but otherwise very different. One of the two agricultural settlements and the State of Israel both had “law”, according to Schwartz, while the highly organized other settlement shared “informal control” with amorphous social groups like the passengers on a cruise ship or the fellow inhabitants of a large apartment building. This sort of perverse classification makes powerful theory about things called “law” impossible.

  • A more general observation is in order. Variation in social life and organization is seldom discontinuous. When social life exhibits continuous variation, theoretical propositions concerning it should be formulated in terms of continuous variables.5757. Cf. Abel (1973bAbel, R. 1973b. “A Comparative Theory of Dispute Institutions in Society.” Law and Society Review 8 (2): 217347.[Crossref], [Google Scholar], 221), arguing generally against the use of dichotomies in the sociology of law. His argument has not received much attention and the use of dichotomies remains ubiquitous. Black’s theory, as he formulates it, is typical in this regard: all social control is either “governmental” or “informal” (Black 1976Black, D. 1976. The Behavior of Law. New YorkAcademic Press. [Google Scholar]). Such an approach stands in the way of powerful theory. Black’s “relational distance” proposition (see text at n 9 above), for example, would be far more powerful if formulated in terms of two continuous variables: “differentiation” in social control and “relational distance”. Many contributions to litigation theory similarly make use of dichotomous variables: Galanter (1974Galanter, M. 1974. “Why the “Haves” Come Out Ahead: Speculations on the Limits of Legal Change.” Law & Society Review 9: 95160.[Crossref], [Google Scholar]) distinguishes between “one shotters” and “repeat players”, Todd (1978Todd, H. 1978. “Litigious Marginals: Character and Disputing in a Bavarian Village.” In The Disputing Process – Law in Ten Societies, edited by L.Nader and H. Todd86121New YorkColumbia University Press. [Google Scholar]) between “insiders” and “marginals”, Merry (1982Merry, S. 1982. “The Social Organization of Mediation in Non-industrial Societies: Implications for Informal Community Justice in America.” In The Politics of Informal Justice (Vol. II), edited by R. Abel1745New YorkAcademic Press. [Google Scholar]) and many others between “mediation” and “adjudication”, Gluckman (1973Gluckman, M. (1955) 1973The Judicial Process Among the Barotse of Northern RhodesiaManchesterManchester University Press. [Google Scholar]) and others between “simplex” and “multiplex” relations, Gulliver (1963Gulliver, P. 1963Social Control in an African SocietyLondonRoutledge & Kegan Paul. [Google Scholar], 299) between “political” and “judicial” dispute processes and so forth. The observations of such authors concerning matters “legal” could easily be formulated in a continuous way, for example, in the case of Todd and Gluckman, by using the continuous variable “relational distance”. Similarly, exchange relationships need not be classified as “contractual” or “non-contractual” but rather, as Macaulay’s (1963Macaulay, S. 1963. “Non-contractual Relations in Business: A Preliminary Study.” American Sociological Review28: 5567.[Crossref], [Web of Science ®], [Google Scholar]) analysis implies, as “more or less contractual”. Such differences of formulation may seem trivial at first sight, but in practice the consequences of discontinuous formulation have got profoundly in the way of theoretical progress in the sociology of law.View all notes Only where nature itself is discontinuous can theory formulated in terms of a discontinuous variable be fruitful. It is in principle wrong for a social science to try to force social life onto the theoretical Procrustes-bed of a discontinuous variable such as “law”.

2.2.3. A non-taxonomic concept of law

The ladder of law has no top and no bottom.

(Bob Dylan)5858. “The lonesome death of Hattie Carroll”.View all notes

The relationship between law and the rest of social control can be formulated in a better way.5959. I first formulated the following idea in Griffiths (1984aGriffiths, J. 1984a. “The Division of Labor in Social Control.” In Toward a General Theory of Social Control, edited by D. Black3770Orlando, FL: Academic Press.[Crossref], [Google Scholar]). There are, of course, earlier suggestions that point in a similar direction (see e.g. Blankenburg 1980Blankenburg, E. 1980. “Recht als Gradualisiertes Konzept” [Law as a Continuous Concept]. Jaarbuch fur Rechtssoziologie und Rechtstheorie 6: 8398. [Google Scholar]; cf. Abel 1973bAbel, R. 1973b. “A Comparative Theory of Dispute Institutions in Society.” Law and Society Review 8 (2): 217347.[Crossref], [Google Scholar], 244–251).View all notes Most taxonomic definitions of law are directly or indirectly concerned with levels of differentiation in social control. Social control above a particular level of differentiation is, by authors in the taxonomic tradition, considered “legal”.6060. The idea that it is differentiation in the sense used here that distinguishes “law” from other social control is latent in the approaches of a number of authors in Box 2, in particular Hoebel (1954Hoebel, E. A. 1954The Law of Primitive ManCambridge, MAHarvard University Press. [Google Scholar], 26–27) but also, less explicitly, Weber, Malinowski, Radcliffe-Brown, Selznick and Black.View all notes But instead of asking ourselves the ultimately unanswerable question, “When is there enough differentiation to be able to speak of law?”, we can simply note that all social control is more or less differentiated. The world does not confront us with distinct sorts – more or less differentiated  of social control but with continuous variation in the degree to which social control is differentiated.

Having taken that simple step, we can then locate the various proposed taxonomic concepts of law on a scale from the imaginary absolute zero point of differentiation in social control (even below the social control observable in everyday interaction in the form of facial expressions, pointed silences and so forth) to the imaginary point of infinite differentiation (with the legislative and adjudicatory institutions of the modern state, for example, toward the upper end of the scale6161. It is of course not the case that all social control for which a state is responsible is equally differentiated, nor that social control by a state is always more differentiated than non-state control, nor that social change is always in the direction of greater differentiation, although the “evolutionary” idea often more or less latent in the sociology of law assume that it is (see for an explicit example, Schwartz and Miller [1964Schwartz, R., and J. Miller1964. “Legal Evolution and Societal Complexity.” American Journal of Sociology 70: 159169.[Crossref], [Web of Science ®], [Google Scholar]]).View all notes). Figure 2 indicates how this can be done.

Figure 2. The relationship between various taxonomic concepts of law on a scale of increasing differentiation in social control.

Figure 2 locates on a common scale the findings and theoretical conclusions of social scientists who have used different taxonomic concepts of law. Despite apparently fundamental conceptual differences, their work can in this way be made commensurable and hence contribute together to the cumulation of knowledge which constitutes scientific progress.

2.3. Conclusion

The most important conclusion of this discussion is that the theoretical object of the sociology of law is not “law” but “social control”. The “legal” aspect of the sociology of law reappears in the form of the continuous variable differentiation.6262. Compare Geiger (1947Geiger, T. 1947Voorstudien zu einer Soziologie des Rechts. Vol. 19, No. 1. AarhusAarhus Universitet, Acta Jutlandica, Aarsskrift. [Google Scholar]). The idea that differentiation is the key to an empirical concept of law is related to Hart’s (1961Hart, H. 1961The Concept of Law. (2nd ed. 1994). OxfordOxford University Press. Page references in this essay are to the 1961 edition. [Google Scholar]) idea that law is to be distinguished from other forms of social control by the fact that law knows not only “primary rules” that regulate social behavior but also “secondary rules” concerning the use of the primary rules (validity, change and adjudication), although Hart seems mistaken in supposing that secondary rules are unique to “law”. See further §3.2.2.View all notes Such a conclusion is not just a matter of terminology. A concept of law that permits continuous variation avoids the distortion of social reality that takes place when, like Schwartz, one struggles to force social control into a dichotomy such as “formal”/“informal” or “legal/not legal”. Such a conception permits us to formulate the “relational distance” proposition as a relationship between two continuous variables: on the one side, the degree of “relational distance”, and on the other side, the degree of differentiation in social control. The reformulated proposition explains not only one (supposedly) “big” difference in social control – that between “law” and the rest – but also all sorts of “smaller” differences, such as that between the social control in Schwartz’ two Israeli settlements, or the different ways contractual relationships are created and adjusted in business communities varying from local merchants in face-to-face relationships to the global world of multinational corporations.6363. Cf. Macaulay (1963Macaulay, S. 1963. “Non-contractual Relations in Business: A Preliminary Study.” American Sociological Review28: 5567.[Crossref], [Web of Science ®], [Google Scholar]1977Macaulay, S. 1977. “Elegant Models, Empirical Pictures and the Complexities of Contract.” Law and Society Review 11: 507529.[Crossref], [Web of Science ®], [Google Scholar]1985Macaulay, S. 1985. “An Empirical View of Contract.” Wisconsin Law Review 1985: 465482. [Google Scholar]1995Macaulay, S. 1995. “Crime and Custom in Business Society.” Journal of Law and Society 22: 248258.[Crossref], [Web of Science ®], [Google Scholar]).View all notes The relational distance proposition, so conceived, is always applicable in all social groups, and not just when and where “law” (according to one or another taxonomic conception) is present. In short, the generality of the relational distance proposition is far greater than when “law” is defined in a taxonomic way. The proposition is more powerful and therefore more vulnerable to a far greater range of empirical tests. The process of testing should quickly reveal difficulties with the current, rather primitive formulation, and hence lead to refinement and reformulation, or to rejection in favor of something better. A non-taxonomic concept of law as more or less differentiated social control will thus stimulate instead of obstructing theoretical progress.6464. If this is convenient, for example, in conversations with or interviews of non-specialists, it is always possible to refer to all social control above a given level of differentiation as “law”. Cf. Moore (1973Moore, S. 1973. “Law and Social Change: The Semi-autonomous Social field as an Appropriate Subject of Study.” Law and Society Review 7: 719746. Reprinted in Moore 1978. London: Routledge & Kegan Paul.[Crossref], [Google Scholar], 745) and Galanter (1981Galanter, M. 1981. “Justice in Many Rooms: Courts, Private Ordering and Indigenous Law.” The Journal of Legal Pluralism and Unofficial Law 13: 147.[Taylor & Francis Online], [Google Scholar], n 16) for such an approach. It may also sometimes be handy to use the term “law” as a pars pro tototo refer to social control at various different levels of differentiation (as in “sociology of law” or “legal pluralism”). There is no objection to a certain amount of this sort of conceptual sloppiness, so long as when one lapses temporarily and for good reasons from conceptual rigor, one remains fully aware of what one is doing and its risks. What is important is that such a folk concept of law be resolutely banned from the professional heart of the discipline: theory.View all notes

I return to the idea that the theoretical object of the sociology of law is social control, and the significance of this idea not only for the sociology of law itself, but more generally for the relationship between sociology of law and sociology, in Section 4.

Social control is highly pluralistic. In all but the most idiosyncratically simple situation, the social groups that exercise it are multifarious, overlapping and competitive with each other. Furthermore, the rules of a group are at any given time not always consistent with one another6565. The group’s secondary rules will generally afford ways of dealing with or eliminating such inconsistencies.View all notes – for example, different state agencies may differ in the rules they seek to enforce, and different groups within a church may have different rules on matters such as birth control. Such differences are typical features of circumstances in which rule change is occurring (which is to say, most of the time). It follows that “law” in the non-taxonomic sense is intrinsically pluralistic and that in the context of empirical theory the expression “legal pluralism”6666. See Griffiths (1986bGriffiths, J. 1986b. “What Is Legal Pluralism?” The Journal of Legal Pluralism and Unofficial Law 18: 155.[Taylor & Francis Online], [Google Scholar]).View all notes – strictly speaking – is redundant.67